Assessing Armour Residential's Performance For Q2 2025

Core Viewpoint - Armour Residential REIT (ARR) experienced a quarter that was largely as expected, with a slight decline in book value and weaker-than-anticipated earnings, primarily due to market fluctuations early in the quarter, although some recovery in June mitigated the impact [1][23]. Book Value and Earnings Performance - ARR reported a minor decrease in book value (BV) of 1.1% during Q2 2025, which was in line with expectations, but it underperformed compared to stronger peers in the agency mREIT sector [11][23]. - The on-balance sheet investment portfolio valuation gain was $16.5 million, closely matching the projected $20.0 million, leading to a BV underperformance of $0.04 per common share [3][23]. - Core earnings/EAD for Q2 2025 were reported at $67.9 million, which was a modest underperformance compared to the estimated $74.0 million, resulting in a variance of $0.071 per share [11][19]. Interest Income and Operational Expenses - Net interest income decreased from $36.3 million in Q1 2025 to $33.1 million in Q2 2025, a decline of $3.2 million, which was below the projected $37.5 million [7][23]. - Operational expenses increased from $13.1 million in Q1 2025 to $14.3 million in Q2 2025, which also contributed to the core earnings underperformance [8][23]. Hedging and Derivative Performance - ARR increased its hedging coverage ratio from 84% to 93% during Q2 2025, which was slightly above the projected range of 85%-90% [4][16]. - The company experienced a derivative valuation loss of $157 million, which was more severe than the anticipated loss of $150 million, contributing to a BV underperformance of $0.09 per common share [4][23]. Sector Comparison and Future Outlook - ARR's performance was weaker compared to peers like AGNC, DX, and NLY, which reported better results in terms of BV and core earnings [23]. - The company is currently rated as a HOLD, with expectations for stabilization in core earnings/EAD in Q3 2025 before any potential rating downgrades [13][24].